In 2012, Spring had come early. Across the Midwest, producers planted their crops ahead of schedule and in favorable conditions. As they parked their planters in the shed for the season, they were left to hope for the best.
What unfolded in that growing season was devastating to millions of acres across the United States. Some of the world’s most productive soils were struck with a drought like nothing in recent history. And while the risks of a disappointing crop are always present, the memories of this drought will live in the minds of agricultural producers for years to come.
Yield risks are to be expected in agriculture. They come with the territory, so to speak. And while the existence of this risk is often discussed, little time and effort is actually spent quantifying the yield risks in agriculture in a practical manner.
By its definition, risk is an unknown future outcome with a distribution of possible results. If we statistically examine the range of yield outcomes, what can we learn about risks in agriculture? One thing we observed is that yield risk can vary greatly within a state. Some counties are much riskier than others. Understanding just how risky yields are in your area can help you build a better financial plan for your farming operation and help you make better financial decisions.
Using historic corn yield data (1950-2012) and accounting for yield-trends over time, we can evaluate yield fluctuation over time. The fluctuation around the trend is one measure of the risks created by droughts, floods, disease, management decisions, and other environmental factors. This fluctuation over time, or variation, was measured by taking the standard deviation of de-trended yields, and dividing it by the mean de-trended yield. This allows us to compare the standard deviation in a common unit.
The first figure shows the average of county yield variations across Indiana, Illinois, and Iowa. Overall, yield variation across counties was lowest in Iowa, at 10.98% of the mean. If one assumes yields follow a normal distribution over time, this means that 68% of yield can be expected +/- 10.98% of the mean yield (there is a 95% probability of a yield occurring at +/- (2* 10.98%) of the mean). Overall, corn yields for Illinois counties are the most risky in the three states.
The second graph shows the distribution of county yield variations across each state. More than 45% of the counties in Iowa have relative standard deviation between 8% and 10% of the average county yield.
Keep in mind the range of these scales. Each state has counties at both ends of the distribution, less than 8% relative standard deviation and more than 16%. The impacts of this are surprising: there are counties in each of the three states that are twice as risky as others.
Risk is not created equally. Producers and those working with producers need to understand that yield variations can come from a number of factors, but certain geographic regions lend themselves to more risk. When farmers talk about risk, it’s important to not use the term generically. Yield risks can be measured and it’s important to know if your farm or your farmer customers can expect yield variations of +/- 8% of average, or greater than +/- 16% of average.
The results are also important when evaluation farmland purchase decisions. Prices should reflect the yield risks in the area. Is the discount/premium for farmland in the area you are considering appropriate given the historical experience with yield risk. Most investors start their analysis with an estimate of the expected yield that they might expect on the farm. These results indicate that one might also want to seriously evaluate how yield variability would influence your investment and financing decisions.
The implications of understanding the quantitative value of yield risk are important to farmers, ag lenders, ag retailer suppliers, and commodity purchasers. Do more than just hope for the best, understand what the worst might look like.
For those interested in learning more, follow the blog and stay connected as we will continue to observe these trends and continue providing the data, charts, and insights.